Shares of Nvidia, AMD, and Intel are experiencing a sharp sell-off on Monday, but this time the trigger is not a disappointing earnings report. Instead, the market is reacting to a more fundamental shift: leading AI executives are questioning whether the industry should continue racing to build ever more powerful models at the current breakneck speed.
Nvidia fell more than 2% in premarket trading, while AMD dropped about 5% and Intel nearly 6%. Nasdaq 100 futures were down 1.72%, reflecting broader tech weakness. The sell-off comes after Anthropic CEO Dario Amodei called on AI companies to slow the pace of model capability advancements over the weekend. OpenAI's Sam Altman and xAI's Elon Musk broadly supported the idea of more restraint.
This introduces a new variable for investors: what happens if the industry intentionally slows frontier AI progress? If development is paced, some data-center compute demand could arrive more slowly than previously anticipated. However, chip orders remain strong, leaving Wall Street to debate whether Monday's decline is a necessary reset or an overreaction.
Brian Jacobsen, chief economist at Annex Wealth Management, told Reuters that "the strongest arguments for caution are those grounded in evidence, not fear," while also warning investors about incumbents protecting their positions. The market is not primarily questioning chip demand, but rather the speed at which future demand will materialize.
The counterargument is that American labs cannot unilaterally stop the global AI race. China continues to invest heavily in models, chips, and sovereign computing infrastructure, while Washington remains focused on preserving technological leadership. Technology analyst Dan Ives noted that Amodei's proposal was an important step toward industry self-regulation, but added that "the reality is China won't slow down anytime soon."
That constraint matters for Nvidia, AMD, and Intel. Pacing model releases does not necessarily mean hyperscalers will cancel data centers or stop purchasing hardware. Safety testing requires significant compute, and inference workloads, sovereign AI projects, and deployment of existing models can continue to absorb chips even if frontier gains arrive more slowly. The market is pricing in slower AI advancement, not confirmed weakness in semiconductor orders.
Just four days before this sell-off, Piper Sandler initiated bullish coverage across the AI semiconductor sector. Analyst David O'Connor called Nvidia the "outright leader in AI compute with 80% market share" and set a $300 price target. He argued that agentic AI has created a step-change in computing demand, with supply likely to remain constrained for another two to three years. Piper also initiated AMD at Overweight with a $600 target, calling it an "Agentic AI Sweetspot" as server CPUs gain share and its Helios GPU platform ramps with OpenAI, Meta, and Anthropic. Intel received a Neutral rating and a $110 target, despite Piper saying agentic AI should support server-CPU demand.
Monday's nearly 6% drop for Intel lands on a company where Wall Street already sees greater execution and valuation risk. The broader sell-off also contrasts with gains in some software stocks—ServiceNow, Adobe, and Workday rose in premarket trading as investors reconsidered how quickly AI might disrupt their businesses.
For investors, the key takeaway is that this sell-off is not about current chip demand, which remains robust, but about the trajectory of future growth. The debate now centers on whether the industry's self-imposed slowdown is realistic given global competitive pressures, and whether the market's reaction is proportionate to the potential impact on AI infrastructure spending.
This article is for informational purposes only and does not constitute financial advice.
